CHARACTERIZATION AND PROPAGATION OF PA'UOHI'IAKA ......Agriculture NEWGERMPLASM Grant for providing...

CHARACTERIZATION AND PROPAGATION OF PA'UOHI'IAKA (JACQUEMONTIA

SANDWICENSIS A. GRAY) FOR POTENTIAL USE AS A HANGING BASKET PLANT

A THESIS SUBMITTED TO THE GRADUATE DIVISION OF THE

UNIVERSITY OF HAWAI‘I AT MĀNOA IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE

IN

TROPICAL PLANT AND SOIL SCIENCES

MAY 2019

By

Darel Kenth S. Antesco

Thesis Committee:

Orville C. Baldos, Chairperson

Teresita D. Amore

Richard Criley

Keywords: Hanging Basket, Native Plants, Pa'uohi'iaka, Morphological Characterization,

Propagation

ii

ACKNOWLEDGEMENTS

I would like to thank the USDA NIFA Hatch Project HAW 080840-H managed by the

College of Tropical Agriculture and Human Resources and the Hawaii Department of

Agriculture NEWGERMPLASM Grant for providing funding for my thesis research. Sincere

thanks to my adviser, Dr. Orville C. Baldos, for the opportunity to work under his research

project. His guidance and mentorship during the course of this study, his sharing valuable

knowledge from experimental set up, writing, statistical analysis and multimedia presentation are

greatly appreciated. I learned so much from his mentorship and am now a better person than I

was on the first day I started as his student. Also, my sincere appreciation to my committee

members, Dr. Teresita Amore and Dr. Richard Criley for their helpful comments and suggestions

in this thesis. Thanks to the Department of Hawaiian Homelands for the collection permits;

Lyon Arboretum and Maui Nui Botanical Gardens for the planting materials. Thank you to Mr.

Craig Okazaki of Magoon Research Facility for providing technical assistance during the

conduct of this study. Thank you to Dr. Robert Paull for giving me access to the weather data

archive of Magoon Research Facility. Thanks to Patrick Thesken and Aleta Corpuz for helping

me in my experiments, and Maria Pamogas Karaan and Smrity Ramavarapu for proofreading my

thesis.

I would also like to thank the University of the Philippines at Los Baños, especially to

Chancellor Fernando Sanchez Jr. PhD, for the recommendation and approval of my study leave

privilege; to my supervisors Prof. Norma Medina, Maria Charito Balladares and Prof. Ryan

Tayobong for allowing me to temporarily vacate my post as a University Research Associate,

and to my co researchers, Archibald Ventura and Nerrisa Cedillo, who assumed my duties and

research responsibilities while I was on study leave.

iii

Lastly, I would like to thank my friends and relatives in the Philippines, especially to my

fiancée and uncle for the moral and financial support in this endeavor. To my late parents, I hope

that they are proud of me.

iv

ABSTRACT

The use of native Hawaiian plants as ornamentals has increased in the last 28 years.

Despite active promotion, efforts to expand selections for horticultural use have been minimal.

Pa‘uohi‘iaka (Jacquemontia sandwicensis A. Gray) is a prostrate-growing, endemic vine

commonly found in coastal areas. In the wild, morphological variation exists but efforts to

collect and characterize these variations for hanging basket use have been limited. To develop

the use of pa‘uohi‘iaka as a hanging basket plant, six accessions were collected, characterized

and assessed for rooting response. Morphological characterization indicated that each accession

has its own unique set of qualitative and quantitative characters. Principal component analysis

identified leaf shape, leaf length, adaxial and abaxial stem color, length of internodes and length

of lateral branches, flower color and number and flowers as important characters that contribute

to the variation of the six accessions. Cluster analysis revealed three distinct groups. Lyon

Arboretum, Puhala Bay and South Point were selected for further evaluation because of their

shorter internodes and lateral branching. Rooting response was associated with high leaf

retention and longer cutting length (i.e. four nodes). Leaf retention was negatively affected by

leaf pubescence. Due to poor rooting and survival of stem cuttings after transplanting, the South

Point accession was dropped for further evaluation.

v

TABLE OF CONTENTS

ACKNOWLEDGEMENTS...……………………………………………………………………..ii

ABSTRACT ................................................................................................................................... iv

TABLE OF CONTENTS ................................................................................................................ v

LIST OF TABLES………………………………………………………………………………..v

LIST OF FIGURES ...................................................................................................................... vii

CHAPTER 1. INTRODUCTION ................................................................................................... 1

CHAPTER 2. MORPHOLOGICAL CHARACTERIZATION AND IDENTIFICATION OF

PA‘UOHI‘IAKA (JACQUEMONTIA SANDWINCENSIS A. GRAY) ACCESSIONS FOR

HANGING BASKET USE ............................................................................................................. 4

Materials and Methods ............................................................................................................. 5

Morphological characterization ....................................................................................... 6

Statistical and Principal Component Analysis ................................................................. 7

Results ....................................................................................................................................... 8

Quantitative Data ............................................................................................................ 10

Principal Component Analysis (PCA) ............................................................................ 12

Cluster Analysis .............................................................................................................. 15

Discussion ............................................................................................................................... 18

CHAPTER 3. EVALUATION OF SINGLE AND FOUR NODE STEM CUTTINGS AS A

PROPAGATION MATERIAL FOR SIX ACCESSIONS OF PA‘UOHI‘IAKA

(JACQUEMONTIA SANDWICENSIS A. GRAY) ........................................................................ 22

Materials and Methods ............................................................................................................ 23

Effect of number of nodes on rooting of accessions ........................................................ 23

Effect of leaf removal and number of nodes on rooting .................................................. 25

Statistical analysis ............................................................................................................ 26

Results ..................................................................................................................................... 26

Effect of number of nodes on rooting of accessions ........................................................ 26

Effect of leaf removal and number of nodes on the rooting of the Ahihi-Kinau accession

....................................................................................................................................................... 37

Discussion ................................................................................................................................ 39

CHAPTER 4. CONCLUSION...................................................................................................... 45

APPENDICES…………………………………………………………………………………...50

LITERATURE CITED ................................................................................................................. 73

vi

LIST OF TABLES

Table 1. 1. Provenance information of pa‘uohi‘iaka accessions used for the................................. 5

Table 1.2.Qualitative morphological characters assessed in six accessions of pa‘uohi‘iaka. ........ 8

Table 1.3. Quantitative data recorded from six accessions of pa‘uohi‘iaka one month after

pruning.. ........................................................................................................................................ 10

Table 1.4. Principal component analysis of the 17 morphological characters.............................. 12

Table 2. 1. Provenance information of pa‘uohi‘iaka

accessions………………………………………………………………………………………..23

Table 2. 2. Average root length of pa‘uohi‘iaka accessions as influenced by propagation dates,

accessions and node number of stem cuttings .............................................................................. 27

Table 2. 3. Average root number of pa‘uohi‘iaka accessions as influenced by propagation dates,


Table 2. 4. Average root number of pa‘uohi‘iaka accessions as influenced by node number of

stem cuttings. ................................................................................................................................ 30

Table 2. 5. Percent rooting of pa‘uohi‘iaka accessions as influenced by propagation dates,


Table 2. 6. Average number of shoots of pa‘uohi‘iaka accessions as influenced by propagation

dates, accessions and node number of stem cuttings. ................................................................... 34

Table 2. 7. Average number of shoots of pa‘uohi‘iaka accessions as influenced by node number

of stem cuttings. ............................................................................................................................ 34

Table 2. 8. Average number of leaves retained by six pa‘uohi‘iaka accessions as influenced by

propagation dates, accessions and node number of stem cuttings.. .............................................. 36

Table 2. 9. Average number of leaves retained by pa‘uohi‘iaka accessions as influenced by node

number of stem cuttings.. .............................................................................................................. 37

vii

LIST OF FIGURES

Figure 1. 1. Leaves, flowers and stems of six pa‘uohi‘iaka accessions ........................................................ 9

Figure 1. 2. Variable factor map of seventeen morphological characters that contributed to the

construction of PC1 and PC2 ...................................................................................................................... 13

Figure 1. 3. Top eight morphological characters of PCA that contributed to the construction of PC1 ...... 14

Figure 1. 4. Top nine morphological characters of PCA that contributed to the construction of PC2. ...... 14

Figure 1. 5. Bi-plot of six pa‘uohi‘iaka accessions ..................................................................................... 15

Figure 1. 6. Individual factor map of the of the six pa‘uohi‘iaka accessions. ............................................ 16

Figure 1. 7. Clustering of the six pa‘uohi‘iaka accessions .......................................................................... 17

Figure 1. 8. Cluster dendrogram of the six pa‘uohi‘iaka accessions ......................................................... 18

Figure 1. 9. Whole plant of six pa‘uohi‘iaka accessions ............................................................................. 20

Figure 2. 1. Pa‘uohi‘iaka (Jacquemontia sandwicensis A. Gray) single and four node ............... 25

Figure 2. 2. Experimental set up of rooting response of six collections of pa‘uohi‘iaka using

single ............................................................................................................................................. 26

Figure 2. 3. Root length of pa‘uohi‘iaka as influenced by propagation dates (S1: March 2018 and

S2: October 2018) and node number of stem cuttings.. ................................................................ 27

Figure 2. 4. Average number of roots of pa‘uohi‘iaka as influenced by propagation dates (S1:

March 2018 and S2: October 2018) and node number of stem cuttings. ..................................... 28

Figure 2. 5. Vigorous rooting of Ahihi-Kinau accession .............................................................. 31

Figure 2. 6. Percent rooting of pa‘uohi‘iaka as influenced by node number of stem cuttings. ... 32

Figure 2. 7. Average number of pa‘uohi‘iaka leaves retained as influenced by propagation dates

(S1: March 2018 and S2: October 2018) and node number of stem cuttings ............................... 35

Figure 2. 8. Average number of roots of Ahihi-Kinau accession stem cuttings as influenced by

number of nodes of stem cuttings and presence and absence of leaves ........................................ 38

Figure 2. 9. Pubescent and glabrous leafed accessions. ............................................................... 41

Figure 2. 10. Leaf retention of the Puhala Bay accession after 21 days in the mist bed. ............ 42

viii

Figure 2. 11. Dead South Point accession after transplanting ..................................................... 43

1

CHAPTER 1

INTRODUCTION

The use of native Hawaiian plants as ornamentals has increased in the last 28 years due to

state laws that require its use in publicly funded landscaping projects (Acts 73 and 236). It has

been encouraged to mitigate the spread of invasive species and to conserve the local biodiversity

(Tamimi, 1999; Ricordi et al., 2014). Promotion of native plants in nurseries and garden centers

can lessen demand and/or replace ornamentals that have escaped cultivation and pose threats to

natural areas (Ruchala, 2002). Despite active promotion, limited plant availability and the lack of

knowledge on the use of native Hawaiian plants continue to be key constraints (Tamimi, 1999

and Ricordi et al., 2014). Studies to develop feasible propagation and production methods are

important to increase the availability of native plants in the nursery trade (Ruchala, 2002).

Pa‘uohi‘iaka (Jacquemontia sandwicensis A. Gray) is a prostrate, endemic vine that

grows in coastal habitats (Wagner et al., 1991). It was formerly classified as J. ovalifolia subsp.

sandwicensis, but molecular and morphological data support it as being a distinct species (Shay

and Drake, 2018; Namoff et al 2010). As an ornamental, pa‘uohi‘iaka is typically used as a

groundcover in landscaping but will also do well in a large pot or hanging basket (Bornhorst and

Rauch, 2003). In the wild, morphological variations of pa‘uohi‘iaka exist. Leaves and stems can

be glabrous to densely tomentose and its flowers maybe pale blue or white (Wagner et al., 1991).

Inflorescence branches and calyces can also vary greatly (Robertson, 1974). Leaf shape can

range from elliptic to suborbicular (Wagner et al., 1991). These variations exist within islands,

populations or even on some individual plants (Robertson, 1974).

Despite the existence of morphological variation, efforts to collect and characterize these

variations for horticultural use have been limited. To increase the variety of pa‘uohi‘iaka in the

2

nursery trade, collection and characterization of wild and cultivated plant material for ornamental

use are key activities. Accurate documentation, characterization and evaluation of germplasm are

essential for effective conservation and use (Biodiversity International, 2007). Characterization

of accessions can identify the ornamental potential of a collection and can also provide

information for future ornamental breeding programs (de Souza et al, 2012).

Due to poor consumer receptivity to native plants (Hooper et al., 2008), studies to

evaluate new uses are needed. Urbanization in Hawaii provides an opportunity to introduce new

uses for pa‘uohi‘iaka. Cultivating these plants in hanging baskets is ideal in an urban setting.

Hanging baskets can fill the need for vertical gardening in small homes that lack landscape

spaces (Starman and Eixmann, 2006).

To increase the availability of pa‘uohi‘iaka selections in the nursery trade, improved

propagation protocols are necessary. Vegetative propagation is the preferred method for

ornamental production. Vegetative propagation maintains uniformity and is a practical solution

to assure a dependable supply of desired genotypes (Zohary, 2001). Although pa‘uohi‘iaka can

be easily propagated from four node stem cuttings (Bornhorst, 1996), propagation using single

node cuttings may be useful to increase limited planting material.

In this thesis, morphological characteristics and rooting response of six pa‘uohi‘iaka

accessions from wild and cultivated sources were assessed. Morphological characterization was

done to determine the identifiable and unique set of characters in each accession The information

provided by morphological characterization was also used to identify and select accessions that

are highly suitable as a hanging basket/container plant. Principal component analysis and cluster

analysis were conducted to determine the important morphological characters that contribute to

the variation and similarity of the accessions.

3

Aside from morphological characterization, the rooting response of the six accessions

were evaluated using four node and single node stem cuttings propagated in two propagating

dates (March and October 2018). The goal was to determine accessions that are most responsive

to rooting and to test the feasibility of using single node stem cutting.

The information generated from the morphological characterization and rooting response

evaluation of the six accession will aid development pa‘uohi‘iaka as a potential hanging basket

plant.

4

CHAPTER 2

MORPHOLOGICAL CHARACTERIZATION AND IDENTIFICATION OF

PA‘UOHI‘IAKA (JACQUEMONTIA SANDWINCENSIS A. GRAY) ACCESSIONS FOR

HANGING BASKET USE

Introduction

The use of native plants in landscaping has been actively promoted in Hawaii for the past

28 years. Despite increased use of native plants in landscaping, a number of challenges still exist

preventing their wide usage. Landscape professionals find it difficult to specify native plants

instead of non-native plant species due to the lack of availability of desired plant species and

sizes as well as the lack of consumer receptivity and customer unfamiliarity with native plants

(Hooper et al., 2016, Ricordi et al., 2014). To increase availability of native Hawaiian plants,

new species and selections must be identified and evaluated for various uses such as hanging

basket plants.

Pa‘uohi‘iaka (Jacquemontia sandwicensis A. Gray) (Convolvulaceae) is a perennial vine

endemic to Hawai‘i. It is commonly found on all main islands at elevations ranging from sea

level to 30.4 m (100 ft) (Wagner et al., 1999). It is a component of coastal vegetation that often

grows with Sida fallax (Shay & Drake, 2018) and is highly salt and wind tolerant (Bezona et al.,

2001). According to Wagner et al (1999), pa‘uohi‘iaka can be glabrous to densely tomentose,

with flowers ranging from white to pale blue. Despite the existence of these variations within the

species, there has been limited efforts to collect and identify selections for naming as cultivars.

As an ornamental plant, pa‘uohi‘iaka has been commonly used as a ground cover for

landscaping. Although it can also be used as a hanging basket or potted plant (Bornhorst &

Rauch, 2003), no selections have been identified for this purpose. In this study, six accessions,

collected from Oahu, Maui and Hawaii Island, were grown in pots and characterized to identify

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=7&cad=rja&uact=8&ved=0ahUKEwjgtO_e3-nWAhVplVQKHYluDu8QFggzMAY&url=https%3A%2F%2Fplants.usda.gov%2Fcore%2Fprofile%3Fsymbol%3DJAOVS&usg=AOvVaw0Y4t8vWjQg9AzdBYJr2Jgy

5

selections for potential use as a hanging basket/container plant. A principal component analysis

and cluster analysis was also done to determine the important morphological characters that

contribute to the variation and similarity of the accessions

Materials and Methods

The study was conducted from October 10, 2017 to February 18, 2018 at Magoon

Research Facility, University of Hawaii at Manoa, Honolulu, USA (Lat: 21.306163, Long: -

157.809243, Elevation: ~48 m above sea level). Pa‘uohi‘iaka accessions were collected as stem

cuttings in situ or from cultivated sources on Oahu, Maui and Hawaii islands (Table 1.1). Stock

plants were established by rooting cuttings on a mist bench in 1:1 by volume mix of perlite and

vermiculite. Rooted cuttings were planted in 15 cm (6 in) plastic pots filled with a 1:1 by volume

mix of coir dust and 1.9 cm (3/4 inch) diameter cinder.

Table 1. 1. Provenance information of pa‘uohi‘iaka accessions used for the characterization

study.

Accession

Name/Provenance

Collection Site Genetic Status

Ahihi-Kinau Maui Nui Botanical Gardens, Maui Wild

Lyon Arboretum* Leeward Community College, Oahu Cultivated, seed bank

accession

McGregor Maui Nui Botanical Gardens, Maui Wild

Puhala Bay Maui Nui Botanical Gardens, Maui Wild

Shidler College* Shidler College Business School,

Oahu

Cultivated

South Point South Point, Hawaii Wild

* unknown provenance

Plants used for morphological characterization were propagated from stock plants on

October 10, 2017. Four to six node cuttings of each accession were treated with Hormex 1

(1000ppm IBA) and inserted vertically in 15 cm (6 in) pots with equal parts of perlite and

6

vermiculite. Cuttings were allowed to root in a mist bench inside a glasshouse for 30 days. Mist

was programmed to turn on for 10 seconds every six minutes. Rooted cuttings were planted

vertically in Deepot Cells D40H (volume: 656 ml; Stuewe and Sons) filled with equal parts of

coconut coir and cinder. Controlled release fertilizer (Nutricote 13-4.8-9.1, Arysta LifeScience)

was also incorporated into the growing media at a rate of 6,992 grams/cubic meter (198

grams/cubic foot). Deepot Cells were placed under full sun and watered twice daily for five

minutes through sprinkler irrigation. Each pot received approximately 220 ml of water daily.

After one month, the plants were potted in 15 cm (6 in) diameter pots using equal parts of

coconut coir and cinder (Appendix Figure 1). Plants were held for another month under the same

outdoor conditions. Since each pot developed only one main stem (Appendix Figure 1.1), plants

were pruned 10 cm (4 inches) from the base to promote lateral branching.

Morphological characterization

Morphological characterization was conducted one month after pruning the plants. Six

plants were used to record a total of 17 qualitative and quantitative traits. Qualitative traits

recorded for each accession were flower color, stem color, leaf pubescence and stem pubescence.

Flower color and stem color were determined using the Royal Horticultural Society 5th edition

(2007) color swatches. Leaf pubescence of the six accessions was categorized either as dense,

medium or absent, while stem pubescence of the six accessions was categorized either as dense,

medium or sparse. Images of the plant and its leaves and flowers were recorded using both a

digital camera (Canon EOS Rebel T7i) and a flatbed scanner (Epson Model EU 88).

Aside from qualitative characters, quantitative characters were also measured from each

plant. Average leaf length, average leaf width, average leaf thickness and average petiole length

were measured from 10 mature leaves of each plant in each accession. Average peduncle length

7

and diameter, and floral diameter were recorded by randomly selecting 10 flowers in the mid

portion of the lateral branches. The average length of internodes was obtained by individually

measuring all the internodes of the longest stem. The average number of lateral branches and the

average length of lateral branches for each plant in each accession were calculated. The average

number of flowers and average number of pre-formed roots (i.e. nodes with pre-formed roots) for

each plant in each accession were also calculated.

Statistical and Principal Component Analysis

Quantitative data were analyzed as a Randomized Complete Block Design with the six

sample plants that served as replicates in Statistix 10 software (Analytical Software) using the

ANOVA function. Tukey HSD was used to separate accession means.

To determine relationships and similarities among the six accessions and to determine the

correlation of morphological characters, principal component analysis and a cluster dendrogram

was generated using the 17 quantitative and qualitative characters. The qualitative data were

transformed by assigning ordinal numbers for each character state. Data were scaled using

prcomp function and were visualized using a combination of ‘FactoMiner’ (Husson et al., 2018)

and ‘factoextra’ (Kassambara and Mundt, 2017: Kasambara, 2018) packages in R studio version

3.4.4 (RStudio, Inc.). Aside from conducting and plotting the Principal Component Analysis, bi-

plot, factor map and individual factor maps were also generated. Contributions of the

morphological characters in the components were also calculated. Principal Component loadings

greater than 0.3 or less than -0.3 were accepted as significant (Peres-Neto et al., 2010; Richman,

1988). To generate the cluster dendrogram, packages ‘dyplyr’, ‘plyr’ and ‘ggplot2’ (Wickham et

al., 2019) were also installed in R studio version 3.4.4.

8

Results

Qualitative Data

Qualitative morphological characters indicate that each accession has its own unique set

and combination of characters. Leaf shape, flower color, stem color and the degree of stem and

leaf pubescence can be used to identify an accession (Table 1.2).

Table 1.2.Qualitative morphological characters assessed in six accessions of pa‘uohi‘iaka.

Accession Leaf

Shape

Leaf

Pubescence

Stem

Pubescence

Stem Color* Flower

Color*

Ahihi-Kinau Obovate None Sparse Purple N77A Violet-

Blue

91C

Lyon

Arboretum

Obovate None Medium Purple N77A White

N155 A

McGregor Ovate Dense Dense Adaxial: Yellow-Green

144D Abaxial:

Purple N77 C

White

N155 A

Puhala Bay Ovate with

undulate

or wavy

leaf

margin

Dense Dense Yellow-Green

144D

Violet-

Blue

91B

Shidler

College

Obovate None Sparse Purple N79B

Violet-

Blue

91B

South Point Ovate Medium Medium Yellow-Green 144D White

N155 A

* Stem and flower color were determined using the Royal Horticultural Society Colour Chart

(2007).

9

Figure 1.1. Leaves, flowers and stems of six pa‘uohi‘iaka accessions; A) Ahihi-Kinau, B) Lyon

Arboretum, C) McGregor, D) Puhala Bay, E) Shidler College and F) South Point.

10

Quantitative Data

At one month after pruning, significant differences were found among accessions for

average leaf length, leaf thickness, petiole length, number of lateral branches, length of

internodes, number of flowers and floral diameter (Table 1.3). No significant differences

between accessions were detected for leaf width (P=0.30), peduncle length (P=0.09), peduncle

diameter (P=0.33), length of lateral branches (P=0.52) and number of pre-formed roots (P=

0.2850).

Table 1.3. Quantitative data recorded from six accessions of pa‘uohi‘iaka one month after

pruning. Means and standard errors presented were rounded off to the nearest tenths. Values with

common letters are not significantly different using Tukey’s HSD pairwise comparison test at P

< 0.05, n=6. Accessions Leaf Length

(cm)

Leaf

Thickness

(mm)

Petiole

Length

(cm)

Number of

Lateral

Branches

Length of

Internodes

(cm)

Flower

Diameter

(mm)

Number of

Flowers

Ahihi-Kinau

4.2±0.1ab 0.42±0.33 a 1.5±0.1a 10.0±0.8 ab 1.7±0.2 ab 11.3±0.4 b 2.8±0.8 ab

Lyon

Arboretum

4.3±0.3 ab 0.33±0.33 b 1.4±0.1 a 13.3±1.8 a 1.4±0.2 b 10.8 b 0.16±0.2 c

McGregor 3.0±0.1c 0.31±0.33 b 1.0 b 9.7±1.2 ab 1.7±0.2 ab 13.0±0.4 b 4.5±0.8 a

Puhala Bay 3.4±0.2bc 0.45±0.33 a 1.4±0.1 ab 12.3±1.6 ab 1.1±0.1 b 11.4±0.4 b 0.7±0.5 bc

Shidler

College

4.7±0.3a 0.32±0.33 ab 1.6±0.1 a 7.0±1.0 b 2.1±0.3 a 14.5±0.4 a 2.2±0.5 abc

South Point 2.9±0.2c 0.33±0.33 b 1.3±0.1 ab 8.2±0.4 ab 1.3±0.2 b 12.0 ±0.3 b 1.7±0.3 bc

Significant differences (P<0.01) in length of leaves were observed among the six

different accessions of pa‘uohi‘iaka. The Shidler College accession exhibited the longest leaf

length (4.7cm) while South Point exhibited the shortest leaf length. Significant differences

(P<0.01) in leaf thickness were also observed between accessions. Puhala Bay possessed the

thickest leaves (0.5mm) while South Point possessed the thinnest leaves (0.3 mm). Average

petiole length among accessions were significantly different (P<0.01). The McGregor accession

11

exhibited the shortest petiole length (1 cm) while the Shidler College accession exhibited the

longest petiole length (1.6 cm). Ahihi-Kinau and Lyon Arboretum exhibited petiole lengths that

were not significantly different from Shidler College. Puhala Bay and South Point accessions

possessed intermediate petiole lengths.

Average number of lateral branches among accessions was significantly different

(P<0.01). Lyon Arboretum exhibited the highest number of lateral branches (13.3) while Shidler

College only produced an average of seven lateral branches. The rest of the accessions possessed

intermediate lateral branch numbers. The average length of internodes of the main stem among

the six accessions was significantly different (P<0.01). The Shidler College accession exhibited

the longest internodes (2.1 cm) while Puhala Bay exhibited the shortest internodes (1.0 cm).

Ahihi-Kinau and McGregor exhibited intermediate internode lengths while Lyon Arboretum and

South Point accessions exhibited similar internode lengths as Puhala Bay.

Average number of flowers between accessions was significantly different (P<0.01).

Lyon Arboretum exhibited the least number of flowers while McGregor exhibited the most

number of flowers. The rest of the accessions exhibited intermediate flower numbers. Average

floral diameter among accessions was significantly different (P<0.01). The Shidler College

accession exhibited the widest floral diameter (14.5 mm) among all accessions.

12

Principal Component Analysis (PCA)

Table 1.4. Principal component analysis of the 17 morphological characters.

PC1 PC2

Eigenvalue 0.90 0.33

Proportion of Variance 0.3961 0.2918

Morphological Characters

Leaf Shape 0.014786 -0.34765

Leaf Length (cm) -0.30864 -0.26303

Leaf Width (cm) -0.29103 -0.28629

Leaf Thickness (mm) -0.01444 -0.05841

Leaf Pubescence 0.284544 0.223899

Adaxial Stem Color -0.35203 -0.1729

Abaxial Stem Color -0.23169 0.300933

Density Stem Pubescence 0.226527 0.089506

Length of Internodes (cm) -0.37612 -0.16403

Number of Lateral Branches 0.17827 -0.16403

Length of Lateral Branches (cm) -0.32344 0.212551

Number preformed roots -0.04846 0.251474

Peduncle Length (cm) -0.0358 0.251474

Peduncle Diameter(mm) 0.249987 -0.29847

Flower Color 0.317332 0.067092

Number of Flowers -0.08901 0.323439

Flower Diameter (mm) -0.23901 0.158758

Table 1.4 shows that 90% of the variation in the morphological characters were explained

by PC1 and only 33% explained by PC2. Since there is a decrease in the variation explained by

PC3 in comparison to PC1 and PC2, only the first two components were reviewed for the

variables (morphological characters) that were used in constructing the variable factor map

(Figure 1.2). Among the 17 morphological characters, five make significant contributions in PC1

and three in PC2. The significant morphological characters in PC1 are length of internodes, stem

color (adaxial), length of lateral branches, flower color and leaf length. In PC2, leaf shape,

number of flowers and stem color (abaxial) are significant.

13

Figure 1.2. Variable factor map of 17 morphological characters that contributed to the

construction of PC1 and PC2. Proximity of arrow points to the perimeter of the circle indicate

strength of correlation between a morphological character and PC1 and PC2. Colors of

morphological characters represent the strength of contribution and importance of each

morphological character (Low: Light green, Medium: Blue and Strong: Red) in the construction

of PC1 and PC2.

Figure 1.2 illustrates the morphological characters that are well represented or important

(long red arrows) and those are not (shorter blue and light green arrows). It also illustrates

morphological characters that are negatively correlated with each other (opposite quadrants).

Among the morphological characters, leaf thickness is the least important in the construction of

the components. Those characters with strong contributions in the components are shown in

Figures 1.3 and 1.4.

14

Figure 1.3. Top eight morphological characters of PCA that contributed to the construction of

PC1. Red dashed line on the graph above indicates the expected average contributions of

morphological characters to the construction of components.

Figure 1. 4. Top nine morphological characters of PCA that contributed to the construction of

PC2. Red dashed line on the graph above indicate the expected average contributions of

morphological characters to the construction of components.

15

The morphological characters that had the most contribution to the construction of PC1

were length of internodes, stem color (adaxial), length of lateral branches, flower color, leaf

length and width, leaf pubescence and peduncle diameter. The morphological characters that had

the most contribution for PC2 were leaf shape, peduncle length, number of flowers, stem color

(abaxial), peduncle diameter, leaf width, density of stem pubescence, leaf length and number of

preformed roots. Characters like leaf thickness, number of lateral branches and flower diameter

did not contribute for construction of these components.

Cluster Analysis

Figure 1.5. Bi-plot of six pa‘uohi‘iaka accessions generated by 17 qualitative and quantitative

morphological characters.

Bi-plot of the six accessions (Figure 1.5) reveal that McGregor is associated with a set of

characters that are not shared by other accessions. Leaf pubescence, density of stem pubescence,

16

peduncle length, number of flowers and preformed roots, stem color (abaxial) makes the

McGregor accession unique from the rest of the accessions and explains why it has its own

cluster (Figure 1.8). South Point and Puhala Bay accessions are closer from one another and

share characters such as number of lateral branches and average peduncle diameter. Ahihi-Kinau,

Shidler College and Lyon Arboretum are also relatively closer from one another compared to the

previous accessions. These accessions were grouped together because of likeness in

morphological characters such as leaf shape, leaf length and width, and adaxial stem color.

Because of these shared characteristics, they belong to the same cluster (Figure 1.6 and Figure

1.7).

Figure 1.6. Individual factor map of the of the six pa‘uohi‘iaka accessions generated by the

qualitative and quantitative morphologic characters. Colors of accessions represent the strength

of each accession in the component.

The individual factor map (Figure 1.6) shows that Shidler College scored high while

Ahihi-Kinau scored the lowest in component 1. McGregor and Puhala Bay accessions scored

high in component 2. Morphological characters near each accession in the Bi-Plot (Figure 1.5) is

what makes it score high in the individual factor map.

17

Figure 1. 7. Clustering of the six pa‘uohi‘iaka accessions generated by seventeen qualitative and

quantitative morphologic characters. Colors represent each clusters.

Figures 1.7 and 1.8 reveal that the six accessions of pa‘uohi‘iaka fall under three major

clusters. The first cluster contained the South Point and Puhala Bay accessions. The second

cluster was the McGregor accession and the third cluster comprised of the Ahihi-Kinau, Shidler

College and Lyon Arboretum accessions.

18

Figure 1.8. Cluster dendrogram of the six pa‘uohi‘iaka accessions generated by seventeen

qualitative and quantitative morphologic characters.

Discussion

The results of this study indicate that each of the six accessions has its own unique set of

qualitative and quantitative characters. No single morphological character significantly affects

the variation of each of the accession but rather a combination of varying set of morphological

characters. Morphological characters like leaf shape, leaf length, adaxial and abaxial stem color,

length of internodes, length of lateral branches, flower color and number of flowers showed

strong correlation and influence on the construction of PC1 and PC2 components. These

characters showed a high level of importance as variables in the variation and strength of each

accession. These characters also strongly influence the clustering pattern of the six accessions

(Figure 1.2, Figure 1.3 and Figure 1.4).

19

The cluster analysis indicated three major clusters (Figure 1.8). Ahihi-Kinau, Lyon

Arboretum and Shidler College formed one cluster. South Point and Puhala Bay accessions

formed the second cluster while McGregor appears as a separate cluster. McGregor possesses

traits that are unique compared to the South Point and Puhala Bay accessions (Figure 1.5). The

distance in dissimilarity suggests that Ahihi-Kinau and Shidler College accessions are likely

more similar to each other than to the Lyon Arboretum accession. Ahihi-Kinau and Shidler

College are more closely related to Lyon Arboretum than South Point and Puhala Bay are to the

McGregor accession cluster.

The bi-plot indicated that the six accessions can be divided into two major groups in

terms of leaf shape, density of stem pubescence and leaf pubescence (Figure 1.5). Puhala Bay,

South Point and McGregor possess an ovate leaf shape with medium to dense stem pubescence

and medium to dense leaf pubescence. Ahihi-Kinau, Lyon Arboretum and Shidler College

possess an obovate leaf shape with medium to sparse stem pubescence and non-pubescent leaves.

Visually, the non-pubescent-leafed accessions have purplish stems while the pubescent-leafed

accessions mostly exhibited green stems. The McGregor accession was an exception since it

exhibited a purplish color on the adaxial part of the stem and green color on the abaxial portion

of the stem (Figure 1.1).

Among the quantitative characters evaluated, the length of internodes and number of

lateral branches were the two ideal traits for selecting accessions suitable for hanging basket use.

In the variable factor map and Bi-plot, these two morphological characters belongs to a group

that are located on opposing quadrants (Figure 1.2 and Figure 1.5). This fits with our criteria for

compact form attributed by shorter internodes and increased number of lateral branches.

20

Among the six accessions, Puhala Bay, Lyon Arboretum and South Point responded well

to pruning by exhibiting a compact form due to shorter internodes and a higher number of

uniformly cascading lateral branches (Table 1.3 and Figure 1.3). In the bi-plot, these accessions

were located close to the vector of number of lateral branches. This means that this particular

trait is being shared by the three selected accessions. In contrast, McGregor, Shidler and Ahihi-

Kinau are located on the opposite end (Figure 1.5). The selected accessions were also located on

the opposite end of the vector of internode length because to the number of lateral branches

(Figure 1.2). Due to these characteristics, Puhala Bay, Lyon Arboretum and South Point were

selected for further evaluation as hanging basket plants.

Figure 1.9. Whole plant of six pa‘uohi‘iaka accessions; A) Ahihi-Kina, B) Lyon Arboretum,

C)McGregor, D) Puhala Bay, E) Shidler College and F) South Point.

Other accessions such as Ahihi-Kinau are vigorous, but do not have a compact

appearance compared to Lyon Arboretum. Ahihi-Kinau could be an ideal landscape ground

cover as it grows vigorously like Shidler College, an accession used in landscaping. Shidler

College did not exhibit compact growth and might not be suitable for use in hanging baskets.

21

Among the six accessions, McGregor has the most number of flowers, does not possess compact

growth and dense foliage under the conditions of this experiment.

This study revealed that within-species variation exists in pa‘uohi‘iaka collections.

Morphological characterization served as a tool for identifying accessions with potential as a

hanging basket plant. It also aided in determining the similarities and relationships between

accessions. Internode length, number of lateral branches, and flower count are the most suitable

characters to use in selecting wild-collected accessions for the purpose of hanging basket

production.

22

CHAPTER 3

EVALUATION OF SINGLE AND FOUR NODE STEM CUTTINGS AS A

PROPAGATION MATERIAL FOR SIX ACCESSIONS OF PA‘UOHI‘IAKA

(JACQUEMONTIA SANDWICENSIS A. GRAY)

Introduction

Stem cuttings are one of the most common propagation methods employed due to low

cost (Hartmann et al., 1997). Vegetative propagation through stem cuttings also produces

uniform planting materials (Maria and Bona, 2010), which is important in ornamental

production. Pa‘uohi‘iaka (Jacquemontia sandwicensis A. Gray) is an endemic, perennial vine

commonly found in coastal areas of the Hawaiian Islands and has the potential to be developed

as a hanging basket plant. It is an easy plant to propagate from cuttings due to the presence of

pre-formed roots on its stems. The recommended length of stem cuttings for pa‘uohi‘iaka should

be 7-10 cm (3 to 4 in) long with two or three nodes per cutting; rooting hormone is not required

(Lilleeng-Rosenberger, 1998). For vegetative propagation to become efficient and reliable,

numerous factors need to be considered, including standardizing the size of stem cuttings and the

rooting response of collections.

Morphological variations exist in pa‘uohi‘iaka (Chapter 1). Each of the six accessions has

its own unique set of morphological characters, but an important consideration is a good ability

to root. It is essential to evaluate the rooting response of these unique accessions to determine

which ones are more responsive to rooting. Multiplying desirable genotypes – selected from

natural variability for different purposes is a major issue for plant germplasm improvement and

maintenance (da Rocha Correa et al., 2011). Developing an effective rooting protocol is crucial

for efficient maintenance of germplasm collections and for generating enough planting material

for evaluation studies. Aside from evaluating the rooting response of each accession, testing the

23

feasibility of single node stem cuttings as a propagation material is also essential. If successful,

single node cuttings would maximize the number of plants propagated at a given time compared

to the current practice. In this study, the rooting response of single node and four node stem

cuttings harvested from each of the six pa‘uohi‘iaka accessions were evaluated in two different

dates (March and October 2018). The rooting response of stem cuttings with or without leaves

was also tested to determine the effect of leaves on root initiation.

Materials and Methods

Effect of number of nodes on rooting of accessions

This study was conducted from March 6 to 27, 2018 and October 2 to 23, 2018 at the

Magoon Research Facility, University of Hawaii at Manoa, Honolulu, USA (Latitude:

21.306616, Longitude -157.809925, Elevation: ~48 m above sea level). The purpose of repeating

the experiment in October 2018 was to validate the results obtained from March 2018. Six

accessions of pa‘uohi‘iaka collected from Oahu, Maui and Hawaii were evaluated in this study

(Table 2.1)

Table 2.1. Provenance information of pa‘uohi‘iaka accessions used for the morphological

characterization study.

Accession

Name/Provenance

Collection Site Genetic Status

Ahihi-Kinau Maui Nui Botanical Gardens, Maui Wild

Lyon Arboretum* Leeward Community College, Oahu Cultivated, seed bank

accession

McGregor Maui Nui Botanical Gardens, Maui Wild

Puhala Bay Maui Nui Botanical Gardens, Maui Wild

Shidler College* Shidler College Business School,

Oahu

Cultivated

South Point South Point, Hawaii Wild

* unknown provenance

24

The stock plants of these accessions were maintained under outdoor irrigated condition

from October 2017 to October 2018 (see Chapter 1). Stock plants were grown in 2 gallon (7.57

liter) plastic pots filled with a 1:1 by volume mix of coconut coir and 1.90 cm (3/4 inch)

diameter cinder. Slow-release fertilizer (Nutricote 13-4.8-9.1, Arysta LifeScience) was also

incorporated into the growing media at 6,992 grams/cubic meter (198 grams/cubic feet). Stock

plants were grown under full sun and watered twice daily for five minutes through irrigation

spray stakes. Each pot received a total of 6.6 liters of water per day. Stem cuttings from the

March and October 2018 studies were gathered from the same group of mother plants.

Single node and four node stem cuttings with pre-formed roots were gathered from the

mid-portion of lateral branches of each accession. Pedicels were removed and the stem cuttings

were planted horizontally in 15 cm (six-inch size) diameter pots with 1:1 by volume perlite and

vermiculite. All pre-formed roots were in contact with the rooting medium at the time of planting

(Figure 2.1). Pots were placed on a mist bench inside a shaded glass house. Misting was set to

operate for 10 seconds every six minutes (Figure 2.2).

The study was laid out in a Split-Split-Plot Design with the two different planting dates

(March and October 2018) serving as the main plot. The six different accessions: Ahihi-Kinau,

Lyon Arboretum, McGregor, Puhala Bay, Shidler College and South Point, served as the

subplot; and the two types of nodal cuttings (four nodes and single node) served as the sub-

subplot. Treatment combinations were replicated four times with each replicate consisting of 10

stem cuttings. Average root length, average root number, average number of leaves retained,

average number of shoots and percent rooting were calculated 21 days after propagation.

Average root length was obtained by calculating the average length of all the roots that initiate

directly from the nodes (Appendix Figure 2.3).

25

Figure 2.1. Pa‘uohi‘iaka (Jacquemontia sandwicensis A. Gray) single (A) and four node

(B) stem cuttings planted in 1:1 by volume of perlite and vermiculite

Effect of leaf removal and number of nodes on rooting

This experiment was conducted from February 14 to March 14, 2018, at the Pope

Laboratory Greenhouse, University of Hawaii at Manoa, USA (Lat: 21.302576, Long: -

157.815111, Elevation: ~30 meters above sea level). The Ahihi-Kinau accession was used for

this experiment. Treatments were laid out in a 2x2 Factorial Completely Randomized Design

with four replicates with each replication consisting of 10 stem cuttings. Factor A was the

number of nodes (single node and four node) of stem cuttings and Factor B was the presence and

absence of leaves. Cuttings were rooted in 15 cm (6 inch) pots filled with a 1:1 by volume mix

of perlite and vermiculite on a mist bench set to open for 20 seconds every two minutes. Average

temperature during the experiment was 22.3°C. Data collected and calculated were the

following; average root length, average root number, and percent rooting were recorded 30 days

after planting.

A B

26

Statistical analysis

Analysis of variance (ANOVA) in Statistix 10 statistical software (Analytical Software)

was used to determine significant treatment effects or interactions. Assumptions for using

ANOVA, e.g. normality and homogeneity of variances, were checked. Significant differences

between treatment means were determined using Tukey’s Honest Significant Difference (HSD)

Test.

Figure 2.2. Experimental set up of rooting response of six collections of pa‘uohi‘iaka using

single and four node stem cuttings

Results

Effect of number of nodes on rooting of accessions

Average root length

ANOVA did not indicate a significant three-way interaction between propagation dates,

accession, and number of nodes (P=0.3136). Significant interactions between propagation dates

and node (P=0.0053), and between propagation dates and accession (P=0.0004) were observed.

In the interaction between propagation dates and node, no differences in root length between the

two dates were observed within single node and within four node cuttings (Figure 2.3). The

27

average root length of four node cuttings was between 6.58 to 7.28 cm. In single node cuttings,

the average root length was between 7.66 and 6.18 cm. No significant differences in average root

length was observed between the four node and single node stem cuttings, except when root

length of four node cuttings recorded in March was compared with root length of single node

cuttings recorded in October.

Figure 2.3. Root length of pa‘uohi‘iaka as influenced by propagation dates (S1: March 2018 and

S2: October 2018) and node number of stem cuttings. Root lengths and standard errors presented

are combined across accessions. Bars with different letters are significantly different using

Tukey’s HSD pairwise comparison test at P<0.05, n=24.

Table 2.2. Average root length of pa‘uohi‘iaka accessions as influenced by propagation dates,

accessions and node number of stem cuttings. Root length and standard errors presented are

combined across node number of stem cuttings. Values that do not have the same letters are

significantly different using Tukey’s HSD pairwise comparison test at P<0.05, n=8.

Accession Average root length (cm)

S1 (March 2018) S2 (October 2018)

Ahihi-Kinau 6.63 bcd 7.04 bcd

Lyon Arboretum 7.61 bc 9.45 ab

McGregor 5.95 cde 4.15 de

Puhala Bay 7.30 bc 3.16 de

Shidler College 9.56 ab 10.62 a

South Point 5.64 cde 5.91 cde

A

ABAB

B

0

1

2

3

4

5

6

7

8

9

S1 S2 S1 S2

Four-nodes Single-node

Aver

age

Root

Len

gth

(cm

)

28

In the interaction between accession and propagation dates, Shidler College, Lyon

Arboretum, Ahihi-Kinau, South Point, and McGregor exhibited similar root lengths between

dates (Table 2.2). The Shidler College accession consistently exhibited the longest average root

length compared to the other accessions. The root length of the Puhala Bay accession was

significantly longer in March 2018 in contrast to October 2018. The South Point and McGregor

accessions consistently exhibited the shortest root length.

Average number of roots

Figure 2.4. Average number of roots of pa‘uohi‘iaka as influenced by propagation dates (S1:

March 2018 and S2: October 2018) and node number of stem cuttings. Number of roots and

standard errors presented are combined across accessions. Bars with different letters are


ANOVA results did not show a significant three-way interaction between propagation

dates, number of nodes, and accession (P=0.6252). Significant interactions between propagation

dates and number of nodes (P=0.0286), propagation dates and accession (P=0.0001) and

accession by nodes (P<0.01) were observed. Results observed in the interaction between

B

A

C

C

0

1

2

3

4

5

6

7

S1 S2 S1 S2


Aver

age

Nu

mb

er o

f R

oots

29

propagation dates and number of nodes indicate that the four node stem cuttings propagated in

October 2018 exhibited a significantly higher number of roots compared to those propagated in

March 2018. Average root numbers of single node stem cuttings between the two dates were

similar, but lower than those observed in cuttings with four nodes.

Table 2. 3. Average root number of pa‘uohi‘iaka accessions as influenced by propagation dates,

accessions and node number of stem cuttings. Number of roots and standard errors presented are



Accessions

Average root number

S1 (March 2018) S2 (October

2018)

Ahihi-Kinau 4.93 b 7.40 a

Lyon Arboretum 3.36 cde 3.18 cde

McGregor 2.44 de 2.42 de

Puhala Bay 2.66 cde 2.91 cde

Shidler College 3.07 cde 4.08 bc

South Point 3.43 bcde 4.60 bc

Results observed in the interaction between propagation dates and accession indicate that

there is a significant increase in the average number of roots for Ahihi-Kinau in October 2018

(Table 2.3). Average number of roots of Ahihi-Kinau increased from 4.93 in March 2018 to 7.40

in October 2018. The rest of the accessions maintained the same average number of roots

between dates.

30

Table 2. 4. Average root number of pa‘uohi‘iaka accessions as influenced by node number of

stem cuttings. Number of roots and standard errors presented are combined across node number

of stem cuttings. Values that do not have the same letters are significantly different using


Accessions Average root number

Four nodes Single node

Ahihi-Kinau 8.60 a 3.72 c

Lyon Arboretum 4.63 bc 1.90 e

McGregor 3.39 cd 1.46 e

Puhala Bay 3.64 c 1.92 e

Shidler College 5.09 b 2.05 e

South Point 5.90 b 2.12 de

In the interaction between accession and nodal number of cuttings, the four node cuttings

of Ahihi-Kinau exhibited the most roots compared to the four node cuttings of all other

accessions (Table 2.4). McGregor exhibited the lowest root numbers among the four node

cuttings of all accessions. Single node cuttings of all accessions except Ahihi-Kinau, exhibited

low root numbers (<3). Figure 2.5 shows the vigorous rooting of Ahihi-Kinau single node and

four node cuttings

31

Figure 2. 5. Vigorous rooting of Ahihi-Kinau accession: A) Four node stem cutting and

B) Single node stem cutting

32

Percent rooting

Figure 2.6. Percent rooting of pa‘uohi‘iaka as influenced by node number of stem cuttings.

Percent rooting and standard errors presented are combined across accessions. Bars with

different letters are significantly different using Tukey’s HSD pairwise comparison test at

P<0.05, n=48.

ANOVA results did not indicate a three-way interaction between propagation dates,

number of nodes, and accession. No significant interaction was also found between the number

of nodes and propagation dates, and between the number of nodes and accession. This allowed

for the pooling of propagation dates and accessions in each nodal cutting. The percent rooting of

four node stem cuttings were significantly higher than single node cuttings (Figure 2.6). A

significant interaction between propagation dates and accession was also observed (P<0.01)

allowing for the pooling of nodal cuttings.

B

A

0

20

40

60

80

100

120

Single-node Four-nodes

Per

cen

t R

oo

tin

g

33

Table 2. 5. Percent rooting of pa‘uohi‘iaka accessions as influenced by propagation dates,

accessions and node number of stem cuttings. Percent rooting and standard errors presented are



Accessions Percent rooting


Ahihi-Kinau 91.25 a 95 a

Lyon Arboretum 86.25 a 82.5 a

McGregor 80 ab 41.25 c

Puhala Bay 85 a 58.75 bc

Shidler College 91.25 a 98.75 a

South Point 77.5 ab 86.75 a

All accessions except McGregor and Puhala consistently exhibited high percent rooting

(>85%) between the two dates, suggesting propagation date effects on the percent rooting of the

McGregor and Puhala Bay accessions (Table 2.5). Both accessions exhibited the lowest percent

rooting in October at 58.75 % for Puhala Bay and 41.25% for McGregor.

Average number of shoots

ANOVA results did not indicate a three-way interaction between propagation dates,

number of nodes, and accession. Significant two-way interactions between propagation dates and

accession (P<0.005), and between the number of nodes and accession (P=0.0154) were

observed.

34

Table 2.6. Average number of shoots of pa‘uohi‘iaka accessions as influenced by propagation

dates, accessions and node number of stem cuttings. Number of shoots and standard errors

presented are combined across node number of stem cuttings. Values that do not have the same

letters are significantly different using Tukey’s HSD pairwise comparison test at P<0.05, n=8.

Accessions Average number of shoots


Ahihi-Kinau 0.46 bcd 0.77 ab

Lyon Arboretum 0.52 bcd 0.79 ab

McGregor 0.31 bcd 0.15 d

Puhala Bay 0.58 abcd 0.13 d

Shidler College 0.35 bcd 0.97 a

South Point 0.41bcd 0.49 b

Results observed in the interaction between accession and propagation dates indicated

that only Shidler College exhibited a significant increase on the average number of shoots in

October 2018 (Table 2.6). The average number of shoots for Shidler College increased from 0.35

in March 2018 to 0.97 in October 2018.

Table 2.7. Average number of shoots of pa‘uohi‘iaka accessions as influenced by node number

of stem cuttings. Number of shoots and standard errors presented are combined across node

number of stem cuttings. Values that do not have the same letters are significantly different using


Accessions Average number of shoots


Ahihi-Kinau 0.92 a 0.31 cde

Lyon Arboretum 0.89 a 0.42 bcde

McGregor 0.31 cde 0.15 de

Puhala Bay 0.43 bcde 0.28 de

Shidler College 0.76 ab 0.56 abcd

South Point 0.67 abc 0.23 de

35

Results observed in the interaction between accession and number of nodes indicate that

the four node cuttings of Ahihi-Kinau, Lyon Arboretum, and South Point exhibited a

significantly higher number of shoots in contrast to one-node cuttings of the same accessions

(Table 2.7). The shoot numbers were not significantly different between four node and single

node stem cuttings of Shidler College, Puhala Bay and McGregor. The shoot numbers from these

accessions were comparable to those observed in the one-node cuttings of Ahihi-Kinau, Lyon

Arboretum and South Point.

Average number of leaves retained

ANOVA indicated no significant three-way interaction between propagation dates,

number of nodes, and accession (P=0.132). Significant interactions between propagation dates

and node (P=0.014), accession and node (P<0.01) and propagation dates and accession

(P<0.01) were observed.

Figure 2.7. Average number of pa‘uohi‘iaka leaves retained as influenced by propagation dates

(S1: March 2018 and S2: October 2018) and node number of stem cuttings. Number of leaves

retained and standard errors presented are combined across accessions. Bars with different letters

are significantly different using Tukey’s HSD pairwise comparison test at P<0.05, n=24.

A

B

C

C

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

S1 S2 S1 S2

Four-nodes Single node

Aver

age

Nu

mb

er o

f L

eaves

Ret

ain

ed

36

Results observed in the interaction between propagation dates and number of nodes

(Figure 2.7) indicated that in general, four node cuttings had a higher number of leaves retained

compared to single node cuttings. In four node cuttings, a significantly lower number of intact

leaves was observed in cuttings propagated in October propagated cuttings than in March. In

March 2018, the average number of leaves retained in four node stem cuttings was 1.52. In

October 2018, the average number of leaves retained was less (0.89). The number of intact

leaves between single node cuttings planted on either date was not significantly different.

Table 2.8. Average number of leaves retained by six pa‘uohi‘iaka accessions as influenced by

propagation dates, accessions and node number of stem cuttings. Number of leaves and standard

errors presented are combined across node number of stem cuttings. Values that do not have the

same letters are significantly different using Tukey’s HSD pairwise comparison test at P<0.05,

n=8.

Accessions Average number of leaves retained


Ahihi-Kinau 1.3 a 0.82 ab

Lyon Arboretum 1.21 a 0.94 ab

McGregor 0.41 bc 0.01 c

Puhala Bay 1.35 a 0.05 c

Shidler College 1.22 a 1.36 a

South Point 0.41 bc 0.127 c

Results observed in the interaction between accession and propagation dates indicate that

in March 2018, South Point and McGregor significantly lost more leaves than the rest of the

accessions tested (Table 2.8). In October 2018, South Point, McGregor and Puhala Bay

significantly lost more leaves than Shidler College, Ahihi-Kinau and Lyon Arboretum. Between

propagation dates, Shidler College, Ahihi-Kinau and Lyon Arboretum exhibited similar leaf

37

numbers. McGregor and South Point also exhibited similar leaf numbers between dates. Puhala

Bay had significantly higher leaf retention in March 2018 in contrast to October 2018.

Table 2.9. Average number of leaves retained by pa‘uohi‘iaka accessions as influenced by node

number of stem cuttings. Number of leaves retained and standard errors presented are combined

across node number of stem cuttings. Values that do not have the same letters are significantly

different using Tukey’s HSD pairwise comparison test at P<0.05, n=8.

No significant differences on the average number of leaves retained were observed in

single node cuttings of all accessions (Table 2.9). Four node stem cuttings of McGregor, Puhala

Bay and South Point exhibited significantly less leaves retained compared to four node stem

cuttings of Ahihi-Kinau, Lyon Arboretum and Shidler College accessions. Four node and single

node stem cuttings of McGregor and South Point accessions similarly exhibited poor leaf

retention.

Effect of leaf removal and number of nodes on the rooting of the Ahihi-Kinau accession

No significant two-way interactions between number of nodes and presence of leaves were

observed for average root length (Appendix Table 2.5) and percent rooting (Appendix Table 2.6).

Both data exhibited significant main effects (i.e. number of nodes and presence/absence of leaves).

Accessions Average number of leaves retained


Ahihi-Kinau 1.78 a 0.33 c

Lyon Arboretum 1.71 a 0.43 bc

McGregor 0.27 c 0.13 c

Puhala Bay 1.05 b 0.35 c

Shidler College 2a 0.58 bc

South Point 0.43 bc 0.10 c

38

Four node stem cuttings exhibited longer root lengths and higher percent rooting compared to

single node stem cuttings (Appendix Figure 2.12 and Appendix Figure 2.15). Cuttings with leaves

had longer roots and higher percent rooting compared to stem cuttings without leaves (Appendix

Figures 2.13 and Appendix Figure 2.14).

Significant two-way interactions were only observed for average root number (P=0.0085).

Four node cuttings with leaves significantly exhibited the highest number of roots (4.82) compared

to other treatments (Figure 2.8). Four node cuttings without leaves, single node cuttings with leaves

and single node cuttings without leaves exhibited low root numbers (<2) and were not significant

from each other.

Figure 2.8. Average number of roots of Ahihi-Kinau accession stem cuttings as influenced by

number of nodes of stem cuttings and presence and absence of leaves. Bars with different letters

are significantly different using Tukey’s HSD pairwise comparison test at P<0.05, n=4.

Overall, four node stem cuttings and stem cuttings with leaves exhibited better rooting

characteristics compared to single node cuttings and stem cuttings without leaves (Appendix

Figure 2.13)

A

BB

B

0

1

2

3

4

5

6

With Leaves No Leaves With Leaf No Leaf


Aver

age

Nu

mb

er o

f R

oots

39

Discussion

Results of the experiments indicate that propagation dates, number of nodes,

accession/leaf type and leaf retention affect rooting success of pa‘uohi‘iaka. The significant

interactions observed between propagation dates and accession, and between propagation dates

and number of nodes, were due to the malfunction of the mist system in October 2018. This

incident resulted in very wet conditions during the rooting period. As a result, root length and

percent rooting of the Puhala Bay accession were negatively impacted. This suggests that Puhala

Bay is sensitive to overwatering (Figure 2.10) and may be an indication that its provenance may

have drier conditions. Hypoxia or oxygen deficiency from excess water may result in a decrease

in root growth in most plants (Friend et al., 1994). In contrast, the wet conditions significantly

increased average number of roots (Table 2.3) of the Ahihi-Kinau accession, indicating that

Ahihi-Kinau favors wetter conditions for root growth. Aside from the propagation dates by

accession interaction, the propagation dates by number of node interaction indicated that root

number of four node cuttings increased under wetter conditions (October 2018).

The number of nodes as well as leaf bearing nodes also appear to influence rooting

success. Single node stem cuttings of all accessions except for Ahihi-Kinau and Shidler College

exhibited poor rooting characteristics compared to four node cuttings of all accessions. Four

node stem cuttings had better rooting characteristics than single node cuttings because single

node stem cuttings contain less leaf and stem tissue. The presence and retention of leaves in

nodal cuttings appear to be an essential parameter in determining not only the ability of

pa‘uohi‘iaka stem cuttings to root but it can also dictate the survival of stem cuttings after

transplanting. Leaves sustain photosynthesis and replenish the carbohydrates and photosynthates

needed to initiate rooting (Tombesi et. al., 2015). There is a positive correlation between high

40

photosynthate content of stem cuttings and rooting because the supply of photosynthates

supports adequate rooting (Hamilton et al., 2002). Without the leaves, stem cuttings may still

produce roots but not as much as those with leaves. The presence of leaves on stem cuttings

influences rooting compounds, such as auxin and co-factors that exert a stimulating effect on

rooting (Maria and Bona, 2010). Leaves also hold auxin (Hartmann et al., 2002) and nutritional

factors leading to adventitious root initiation (Jarvis, 1986).

The degree of pubescence on the leaves observed in each accession also appear to have

an impact on rooting of pa‘uohi‘iaka. In general, accessions with glabrous leaves (Figure 2.9)

tend to have significantly higher average number of shoots and number of leaves retained. This

happened despite the irrigation malfunction in our experiment, suggesting that Shidler College

(glabrous leaf type) responded favorably to wetter conditions. Accessions with medium to

densely pubescent leaves (Figure 2.16) tend to have low leaf retention and fewer roots. These

observations were further supported when data were re-analyzed into pubescent and glabrous

types (Appendix Figures 2.17 to 2.19 and Appendix Tables 2.8 to 2.10). Under mist conditions,

the accessions with pubescent leaves turned yellowish and defoliated faster than glabrous types.

Survival of rooted cuttings from pubescent accessions also appeared to be low after transplanting

because of poor leaf retention (Darel Kenth Antesco, personal observation).

41

Figure 2. 9. Pubescent (top) and glabrous (bottom) leafed accessions of pa‘uohi‘iaka evaluated

in the nodal cutting study.

42

Figure 2.10. Leaf retention of the Puhala Bay accession after 21 days in the mist bed: A) four

node stem cuttings in March 2018, B) single node stem cuttings in March 2018, C) four node

stem cuttings in October 2018 and D) single node stem cuttings in October 2018.

43

Cuttings of many plant species can develop roots but do not survive for a long time after

rooting (Hartmann and Kester 1983). This is possibly caused by the inability to recover after

transplanting or the failure to adapt to the field environment (Berhe and Negash 1998). In this

study, we observed the difficulty of establishing rooted cuttings of the South Point accession

after transplanting (Figure 2.11). Although South Point successfully developed roots and shoots

under mist conditions, the number of leaves retained in the stem cuttings were low compared to

other accessions (Table 2.9). At transplanting, most rooted cuttings of South Point have lost their

leaves (Darel Kenth Antesco, personal observation). The depletion of photosynthates stored in

the cuttings even before shoots were able to photosynthesize was probably the reason why

majority of the cuttings died after potting. Due to the difficulty of achieving enough number of

plants, the accession was dropped for hanging basket evaluation.

Figure 2.11. Mortality of South Point rooted cuttings after transplanting

44

Among the six accessions, Ahihi-Kinau and Shidler College can be considered as the

most rooting responsive. These are also the only accessions that successfully rooted using single

node stem cuttings. Despite the overwatering incident in October 2018, these accessions

exhibited enhanced rooting response by producing significantly higher number of roots (Ahihi-

Kinau) or shoots (Shidler College) compared to cuttings propagated in March 2018. The low

average number of leaves retained in the McGregor, Puhala Bay and South Point accessions

indicated sensitivity of pubescent types to wet conditions. These accessions average less than 1

leaf retained in the four node stem cuttings (Table 2.9). In contrast, glabrous accessions like

Ahihi-Kinau, Lyon Arboretum and Shidler College have greater tolerance to wet conditions.

Ahihi-Kinau significantly increased average number of roots while Shidler College significantly

increased average number of shoots in the second propagation dates (October 2018). In the first

propagation date (March 2018), Puhala Bay (pubescent leaf type) responded well to the misting

irrigation interval of 10 seconds every six minutes. Other pubescent leaf types like McGregor

and South Point accessions responded poorly under this condition, suggesting that these

accessions might require a different condition of propagation, i.e. not under mist or less watering

intervals. The McGregor accession can be considered as the weakest among the six accessions

evaluated for rooting response under mist conditions. It recorded low rooting response for all

parameters measured in two propagation dates. Thus, further study focusing on the rooting

response of these three accessions on drier or less frequent watering interval is needed to

improve the efficiency of maintaining these accessions.

45

CHAPTER 4

CONCLUSION

Findings in the morphological characterization and rooting response studies generated

useful information for identifying pa‘uohi‘iaka hanging basket selections and developing

production protocols. Documentation of the traits for each accession provided a snapshot of the

morphological variation in pa‘uohi‘iaka. Morphological characters that significantly contributed

to the variation of accessions were leaf shape, leaf length, adaxial and abaxial stem color, length

of internodes, length of lateral branches, flower color and number of flowers. Among the

morphological characters examined, shorter internodes and a higher lateral branch counts were

the two most important characters to consider for identifying the suitability of selections for

hanging basket use. Cluster analysis of the six accessions revealed three distinct groupings.

Accessions with the same leaf shape, stem color and leaf pubescence (i.e. glabrous or pubescent)

tend to group together in the same cluster. Identification of these important characters and

relationships between accessions provides relevant information that can be used for developing

new cultivars in future breeding programs.

The rooting response study revealed that four node stem cuttings were better propagation

materials compared to single node stem cuttings. Four node stem cuttings possessed more

preformed roots as well as more leaf and stem tissue to support root growth. Leaf retention and

rooting of the six accessions also appeared to be influenced by leaf pubescence. The glabrous

leafed accessions (Ahihi-Kinau, Shidler College and Lyon Arboretum) rooted well because of a

significantly higher number of leaves retained. In contrast, pubescent leafed accessions

(McGregor, Puhala Bay and South Point) rooted poorly due to rotting of most leaves. The wet

conditions in the mist bench caused the leaves of the pubescent leafed accessions to defoliate,

46

indicating that rooting of pubescent leafed accessions may require drier growing conditions or

lower frequency of misting. Leaves support the root initiation of stem cuttings through a

continuous supply of photosynthates. It also ensures survival of rooted cuttings after

transplanting. Due to poor rooting and survival of South Point, only two accessions, Lyon

Arboretum and Puhala Bay), were advanced to the hanging basket trials. The trials evaluating the

response of the two accessions to different frequencies of manual pinching are currently ongoing.

Overall, this thesis indicated that inherent variation in native plants can be used for

developing ornamental selections for particular uses. It also showed that propagation protocols

may vary depending on characteristics of an accession. Further collection is needed to increase

and assess the diversity of pa‘uohi‘iaka for future evaluation and breeding programs.

47

APPENDICES

Appendix Table 1.1 .Analysis of variance (ANOVA) of average leaf length (cm) of six

pa‘uohi‘iaka accessions one month after pruning.

Source df SS MS F P

Samples 5 1.0674 0.21348

Accession 5 17.1139 3.42278 13.35 0

Error 25 6.4105 0.25642

Total 35 24.5919

Appendix Table 1. 2. Analysis of variance (ANOVA) of average leaf width (cm) of six


Source df SS MS F P

Samples 5 3.0142 0.60284

Accession 5 4.1742 0.83483 1.27 0.3094

Error 25 16.4846 0.65938

Total 35 23.673

Appendix Table 1. 3. Analysis of variance (ANOVA) of average leaf thickness (mm) of six


Source df SS MS F P

Samples 5 0.00865 0.00173

Accession 5 0.17064 0.03413 8.17 0.0001

Error 23 0.09608 0.00418

Total 33

Appendix Table 1.4. Analysis of variance (ANOVA) of average peduncle length (cm) of six


Source df SS MS F P

Samples 5 2.7523 0.55047

Accession 5 8.637 1.72741 2.21 0.0977

Error 18 14.0378 0.77988

Total 28

48

Appendix Table 1.5. Analysis of variance (ANOVA) of average peduncle diameter (mm) of six


Source df SS MS F P

Samples 5 0.11927 0.02385

Accession 5 0.07221 0.01444 1.24 0.3369

Error 16 0.18654 0.01166

Total 26

Appendix Table 1.6. Analysis of variance (ANOVA) of average petiole length (cm) of six


Source df SS MS F P

Samples 5 0.06742 0.01348

Accession 5 0.98982 0.19796 4.82 0.0032

Error 25 1.02618 0.04105

Total 35 2.08342

Appendix Table 1.7. Analysis of variance (ANOVA) of average length of internodes (cm) of six


Source df SS MS F P

Samples 5 1.41272 0.28254

Accession 5 4.23672 0.84734 5.66 0.0013

Error 25 3.74108 0.14964

Total 35 9.39052

Appendix Table 1.8. Analysis of Variance (ANOVA) of number of lateral branches of six


Source df SS MS F P

Samples 5 5.222 1.0444

Accession 5 174.222 34.8444 3.25 0.0215

Error 25 268.444 10.7378

Total 35 447.889

49

Appendix Table 1.9. Analysis of variance (ANOVA) of average length of lateral branches of six


Source df SS MS F P

Samples 5 567.78 113.556

Accession 5 553.6 110.72 2.59 0.0521

Error 24 1026.16 42.757

Total 34

Appendix Table 1.10. Analysis of variance (ANOVA) of average number of flowers of six


Source df SS MS F P

Samples 5 13.333 2.6667

Accession 5 73.333 14.6667 7.75 0.0002

Error 25 47.333 1.8933

Total 35 134

Appendix Table 1.11. Analysis of variance (ANOVA) of average floral diameter of six


Source df SS MS F P

Samples 5 4.5194 0.90388

Accession 5 36.5553 7.31107 11.03 0.0001

Error 15 9.9398 0.66266

Total 25

Appendix Table 1.12. Analysis of variance (ANOVA) of average number of preformed roots of

six pa‘uohi‘iaka accessions one month after pruning.

Source df SS MS F P

Samples 5 403.14 80.628

Accession 5 748.47 149.694 1.33 0.285

Error 25 2819.36 112.774

Total 35 3970.97

50

Appendix Table 2.1. Analysis of variance (ANOVA) for root length (cm) of six pa‘uohi‘iaka

accessions propagated from one-node and four node cuttings in March and October 2018

Source df SS MS F P

Rep (A) 3 8.214 2.738

Propagation dates (B) 1 3.682 3.6817 6.49 0.0841

Error A*B 3 1.701 0.5671

Accession (C) 5 324.548 64.9095 21.32 0

B*C 5 96.824 19.3647 6.36 0.0004

Error A*B*C 30 91.318 3.0439

Nodes (D) 1 0.007 0.0067 0 0.9644

B*D 1 29.018 29.018 8.81 0.0053

C*D 5 13.193 2.6386 0.8 0.5565

B*C*D 5 20.326 4.0651 1.23 0.3136

Error A*B*C*D 36 118.608 3.2947

Total 95 707.437

Appendix Table 2.2. Analysis of variance (ANOVA) for number of roots of six pa‘uohi‘iaka


Source df SS MS F P

Rep (A) 3 12.122 4.041


Error A*B 3 2.375 0.792

Accession (C) 5 141.174 28.235 55.9 0

B*C 5 19.604 3.921 7.76 0.0001

Error A*B*C 30 15.152 0.505

Nodes (D) 1 217.262 217.262 347.85 0

B*D 1 3.249 3.249 5.2 0.0286

C*D 5 28.057 5.611 8.98 0

B*C*D 5 2.194 0.439 0.7 0.6252

Error A*B*C*D 36 22.485 0.625

Total 95 478.433

51

Appendix Table 2.3. Analysis of variance (ANOVA) for percent rooting of six pa‘uohi‘iaka


Source df SS MS F P

Rep (A) 3 294.8 98.3


Error A*B 3 236.5 78.8

Accession (C) 5 13655.2 2731 18.92 0

B*C 5 7821.9 1564.4 10.84 0

Error A*B*C 30 4331.2 144.4

Nodes (D) 1 13776 13776 66.01 0

B*D 1 1 1 0 0.9441

C*D 5 2430.2 486 2.33 0.0624

B*C*D 5 530.2 106 0.51 0.7682

Error A*B*C*D 36 7512.5 208.7

Total 95 52174

Appendix Table 2. 4. Analysis of variance (ANOVA) table for number of shoots of six

pa‘uohi‘iaka accessions propagated from one-node and four node cuttings in March and October

2018

Source df SS MS F P

Rep (A) 3 0.096 0.03199

Dates (B) 1 0.3049 0.30488 2.39 0.22

Error A*B 3 0.383 0.12766

Accession (C) 5 2.5506 0.51013 10.85 0

B*C 5 2.8819 0.57638 12.26 0

Error A*B*C 30 1.4105 0.04702

Nodes (D) 1 3.0353 3.03526 78.42 0

B*D 1 0.0982 0.09818 2.54 0.12

C*D 5 0.6335 0.12671 3.27 0.0154

B*C*D 5 0.2209 0.04418 1.14 0.3567

Error A*B*C*D 36 1.3934 0.03871

Total 95 13.0081

52

Appendix Table 2.5. Analysis of variance (ANOVA) for number of leaves retained of six

pa‘uohi‘iaka accessions propagated from one-node and four node cuttings in March and October

2018

Source df SS MS F P

Rep (A) 3 0.6511 0.217

Dates (B) 1 4.4376 4.4376 54.95 0.0051

Error A*B 3 0.2423 0.0808

Accession (C) 5 16.434 3.2868 27.37 0

B*C 5 4.5177 0.9035 7.52 0.0001

Error A*B*C 30 3.603 0.1201

Nodes (D) 1 18.8151 18.8151 131.14 0

B*D 1 0.9401 0.9401 6.55 0.0148

C*D 5 6.5881 1.3176 9.18 0

B*C*D 5 1.3116 0.2623 1.83 0.132

Error A*B*C*D 36 5.1651 0.1435

Total 95 62.7055

Sub Study: Effect of leaf removal and number of nodes on the rooting of pa‘uohi‘iaka

Appendix Table 2.6. Analysis of variance (ANOVA) table for root length of pa‘uohi‘iaka

propagated from one-node with leaves and no leaves and four node with leaves and no leaves

stem cuttings.

Source df SS MS F P

Number of Nodes 1 8.8283 8.8283 16.3 0.0016

Leaves 1 21.9141 21.9141 40.46 0

Number of Nodes * Leaves 1 2.1572 2.1572 3.98 0.0692

Error 12 6.4987 0.5416

Total 15 39.3984

Appendix Table 2 7.Analysis of variance (ANOVA) table for percent rooting of pa‘uohi‘iaka

propagated from stem cuttings with one-node with leaves and no leaves and four node with

leaves and no leaves

Source df SS MS F P

Number of Nodes 1 4225 4225 13.52 0.0032

Leaves 1 4900 4900 15.68 0.0019

Number of Nodes *

Leaves 1 25 25 0.08 0.7821

Error 12 3750 312.5

Total 15 12900

53

Appendix Table 2.8. Analysis of variance (ANOVA) for number of roots of pa‘uohi‘iaka

propagated from stem cuttings with one node with leaves or no leaves and four nodes with leaves

or no leaves .

Source df SS MS F P

Number of Nodes 1 13.286 13.286 17.73 0.0012

Leaves 1 19.847 19.847 26.49 0.0002

Number of Nodes * Leaves 1 7.3984 7.3984 9.87 0.0085

Error 12 8.9917 0.7493

Total 15 49.5232

Combined data of glabrous and pubescent leaf type accessions

Appendix Table 2. 9. Analysis of variance (ANOVA) table for number of leaves retained of

glabrous and pubescent pa‘uohi‘iaka accessions propagated from one-node and four node

cuttings in March and October 2018.

Source df SS MS F P

Rep 3 0.6511 0.217

Dates 1 4.4376 4.4376 54.95 0.0051

Error Rep*Dates 3 0.2423 0.0808

Leaf Type 1 13.5901 13.5901 78.83 0.0001

Dates*Leaf Type 1 1.2421 1.2421 7.2 0.0363

Error Rep*Dates*Leaf Type 6 1.0344 0.1724

Nodes 1 18.8151 18.8151 130.6 0

Dates*Nodes 1 0.9401 0.9401 6.53 0.0253

Leaf Type*Nodes 1 5.8707 5.8707 40.75 0

Dates*Leaf Type*Nodes 1 0.1785 0.1785 1.24 0.2874

Error Rep*Dates*Leaf Type*Nodes 12 1.7289 0.1441

Error 64 13.9745 0.2184

Total 95 62.7055

54

Appendix Table 2. 10. Analysis of variance (ANOVA) table for number of roots of of glabrous

and pubescent pa‘uohi‘iaka accessions propagated from one-node and four node cuttings in

March and October 2018.

Source df SS MS F P

Rep 3 12.122 4.041

Dates 1 14.758 14.758 18.64 0.0229


Leaf Type 1 38.153 38.153 191.97 0

Dates*Leaf Type 1 2.331 2.331 11.73 0.0141

Error Rep*Dates*Leaf Type 6 1.192 0.199

Nodes 1 217.262 217.262 364.79 0

Dates*Nodes 1 3.249 3.249 5.45 0.0377

Leaf Type*Nodes 1 6.923 6.923 11.62 0.0052

Dates*Leaf Type*Nodes 1 0.01 0.01 0.02 0.899

Error Rep*Dates*Leaf Type*Nodes 12 7.147 0.596

Error 64 172.91 2.702

Total 95 478.433

Appendix Table 2. 11. Analysis of variance (ANOVA) table for root length of glabrous and

pubescent pa‘uohi‘iaka accessions propagated from one-node and four node cuttings in March

and October 2018.

Source df SS MS F P

Rep 3 8.214 2.738

Dates 1 3.682 3.682 6.49 0.0841


Leaf 1 235.188 235.188 111.83 0

Dates*Leaf 1 53.76 53.76 25.56 0.0023

Error Rep*Dates*Leaf 6 12.618 2.103

Nodes 1 0.007 0.007 0 0.9728

Dates*Nodes 1 29.018 29.018 5.29 0.0402

Leaf*Nodes 1 0.465 0.465 0.08 0.7759

Dates*Leaf*Nodes 1 14.369 14.369 2.62 0.1314

Error Rep*Dates*Leaf*Nodes 12 65.792 5.483

Error 64 282.624 4.416

Total 95 707.437

55

Appendix Figure 1.1. Potted pa‘uohi‘iaka plants before pruning to four inches from the base

Appendix Figure 1.2. Monthly mean, maximum and minimum temperature at Magoon Research

Facility from October 2017 to March 2018.

25.2523.94

21.8322.88 22.3

32.36 31.64

29.07 29.34 30.14

17.315.87

13.88

16.8915.44

0

5

10

15

20

25

30

35

Oct-17 Nov-17 Dec-17 Jan-18 Feb-18

Temperature (Celsius)

Monthly Mean Monthly Maximum Monthly Minimum

56

Appendix Figure 1.3. Monthly mean, maximum and minimum relative humidity at Magoon

Research Facility from October 2017 to March 2018

Appendix Figure 1.4.Cumulative monthly precipitation (mm) at Magoon Research Facility from

October 2017 to March 2018.

98.5 96.599.5 97.6 98.7

47.7 47.443.7

51.8

46.4

75.75 75.06 76.95 78.5581.59

0

20

40

60

80

100

120

Oct-17 Nov-17 Dec-17 Jan-18 Feb-18

Rel

ati

ve

Hu

mid

ity

Monthly Maximum Monthly Minimum Monthly Mean

0

50

100

150

200

250

300

Oct-17 Nov-17 Dec-17 Jan-18 Feb-18 Mar-18

Cumulative Precipitation (mm)

57

Appendix Figure 1.5. Day length graph for Honolulu, USA from October 2017 to March 2018.

Source: www.timeanddate.com/sun/usa/honolulu?month=4&year=2017.

58

Appendix Figure 2.1. Maximum, mean and minimum monthly temperature at the Magoon

Research Facility, Honolulu, Hawaii (~48 m above sea level) from October 2017 to October

2018.

Appendix Figure 2.2. Maximum, mean and minimum monthly relative humidity at the Magoon

Research Facility, Honolulu, Hawaii (~48 m above sea level) from October 2017 to October

2018.

32.36 31.6429.07 29.34 30.14 30.27 29.74 30.95

32.72 31.84 32.33 33.26 33.68

25.2523.94

21.83 22.88 22.3 22.323.72 24.3 25.53 26.29 26.56 26.31 25.82

17.315.87

13.8816.89

15.44 16.37 15.3716.8

20.17 20.5822.44

20.02 18.84

0

5

10

15

20

25

30

35

40

Tem

per

atu

re (

Cel

siu

s)

Max Temp Mean Monthly Temp Min Temp

98.5 96.599.5 97.6 98.7 99 97.6 97.2 97.6 96.3 97.6 98.6 98.9

75.75 75.06 76.96 78.5581.59

77.31 78.8 76.07 73.51 73.95 75 77.73 78.92

47.7 47.443.7

51.846.4

4145.7

48.9

40.6

55.22

47.354.3 53.2

0

20

40

60

80

100

120

Rel

ati

ve

Hu

mid

ity

Max RH Mean Monthly RH Min RH

59

Appendix Figure 2.3. Rooting at the nodes of pa‘uohi‘iaka stem cuttings.

60

Appendix Figure 2.4. Rooting of four node stem cuttings of six pa‘uohi‘iaka (Jacquemontia

sandwicensis A. Gray) accessions propagated in March 2018: A) Ahihi-Kinau, B) Lyon

Arboretum, C) McGregor, D) Puhala Bay, E) Shidler College, and F) South Point.

61

Appendix Figure 2. 5. Rooting of single node stem cuttings of six pa‘uohi‘iaka (Jacquemontia

sandwicensis A. Gray) accessions propagated in March 2018: A)Ahihi-Kinau, B) Lyon

Arboretum, C) McGregor, P) Puhala Bay, E) Shidler College, and F) South Point

62

Appendix Figure 2. 6. Rooting of four node stem cuttings of six pa‘uohi‘iaka (Jacquemontia

sandwicensis A. Gray) accessions propagated in October 2018: A) Ahihi-Kinau, B) Lyon


63

Appendix Figure 2. 7. Rooting of single node stem cuttings of six pa‘uohi‘iaka (Jacquemontia

sandwicensis A. Gray) accessions propagated in October 2018: A)Ahihi-Kinau, B) Lyon


64

Appendix Figure 2.8. Leaf retention of four node stem cuttings of six pa‘uohi‘iaka

(Jacquemontia sandwicensis A. Gray) accessions 21 days after propagation under the mist

bench. Cuttings were propagated in March 2018. A) Ahihi-Kinau, B) Lyon Arboretum, C)

McGregor, D) Puhala Bay, E) Shidler College, and F) South Point

65

Appendix Figure 2. 9. Leaf retention of the six accessions of pa‘uohi‘iaka (Jacquemontia

sandwicensis A. Gray) four node stem cuttings for second propagation date (October 2018).

66

Appendix Figure 2.10. Leaf retention of the six accessions of pa‘uohi‘iaka (Jacquemontia

sandwicensis A. Gray) single node stem cuttings from first propagation dates (March 2018) of

propagation. A) Ahihi- Kinau, B) Lyon Arboretum, C) McGregor, D) Puhala Bay, E) Shidler

College, and F) South Point.

67

Appendix Figure 2.11. Leaf retention of the six accessions of pa‘uohi‘iaka (Jacquemontia

sandwicensis A. Gray) single node stem cuttings from first (March 2018) and second

propagation dates (October 2018) of propagation.

68

Appendix Figure 2.12. Average root length (cm) of Ahihi-Kinau stem cuttings as influenced by

number of nodes. Root length and standard errors presented are combined across stem cuttings

with or without leaves. Bars that are not the same letters are significantly different using Tukey’s

HSD pairwise comparison test at P<0.05, n=8.

Appendix Figure 2.13. Average root length (cm) of Ahihi-Kinau accession single node stem

cuttings as influenced by presence and absence of leaves. Root length and standard errors

presented are combined across stem cutting lengths. Bars that are not the same letters are


A

B

0

0.5

1

1.5

2

2.5

3

3.5

4


Av

era

ge

Ro

ot

Len

gth

(cm

)

A

B

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

With Leaves No Leaves

Av

era

ge

Root

Len

gth

(cm

)

69

Appendix Figure 2. 14. Percent rooting of Ahihi-Kinau stem cuttings as influenced by number of

nodes. Percent rooting and standard errors presented are combined across stem cuttings with or

without leaves. Bars that are not the same letters are significantly different using Tukey’s HSD

pairwise comparison test at P<0.05, n=8.

Appendix Figure 2.15. Percent rooting of Ahihi-Kinau accession single node stem cuttings as

influenced by and presence and absence of leaves. Percent rooting and standard errors presented

are combined across stem cutting lengths. Bars that are not the same letters are significantly

different using Tukey’s HSD pairwise comparison test at P<0.05, n=8.

A

B

0

10

20

30

40

50

60

70

80

90

100


Per

cen

t R

oo

tin

g

A

B

0

10

20

30

40

50

60

70

80

90

100

With Leaves No Leaves

Per

cen

t R

ooti

ng

70

Appendix Figure 2.16. Rooting of Ahihi-Kinau stem cuttings: A) four nodes and no leaves, B)

Four nodes with leaves, C) single node with leaf and D) single node without leaf.

71

Appendix Figure 2.17. Combined data on the average root length of pa‘uohi‘iaka as influenced

by propagation dates (S1: March 2018 and S2: October 2018) and leaf type of accessions.

Number of leaves retained and standard errors presented are combined across leaf type of

accessions. Bars that are not the same letters are significantly different using Tukey’s HSD

pairwise comparison test at P<0.05, n=24.

Appendix Figure 2.18. Combined data on the average number of leaves retained of pa‘uohi‘iaka

as influenced by propagation dates (S1: March 2018 and S2: October 2018) and leaf type of

accessions. Number of leaves retained and standard errors presented are combined across leaf

type of accessions. Bars that are not the same letters are significantly different using Tukey’s


A

B

C

A

0

1

2

3

4

5

6

7

8

9

10

Glabrous Pubescent Pubescent Glabrous

Season 1 Season 2

Av

era

ge

Ro

ot

Len

gth

(cm

)

A

B

AB

C

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

Glabrous Pubescent Glabrous Pubescent

Season 1 Season 2

Av

erage

Nu

mb

er o

f L

eaves

Ret

ain

ed

72

Appendix Figure 2.19. Combined data on the average number of roots of pa‘uohi‘iaka as

influenced by propagation dates (S1: March 2018 and S2: October 2018) and leaf type of

accessions. Number of leaves retained and standard errors presented are combined across leaf

type of accessions. Bars that are not the same letters are significantly different using Tukey’s


B

C

A

BC

0

1

2

3

4

5

6

Glabrous Pubescent Glabrous Pubescent

Season 1 Season 2

Av

era

ge

Nu

mb

er o

f R

oo

ts

73

LITERATURE CITED

Bezona N., Hensley D.,Yogi J., Tavares J., Rauch F., Iwata R., Kellison M. and Wong M. 2001.

Salt and wind tolerance of landscape plants for Hawaii. CTAHR Cooperative Extension.

Available: dlnr.hawaii.gov/occl/files/2013/08/001-salt-tolerance.pdf [Accessed January 7,

2019].

Bioversity International. 2007. Guidelines for the development of crop descriptor lists, (13), 84.

Available:www.bioversityinternational.org/fileadmin/_migrated/uploads/tx_news/Developing_crop_descriptor_lists_1226.pdf [Accessed January 7, 2018].

Bornhorst, Heidi L. 1996. Growing native Hawaiian plants: a how-to guide for the

gardener. Honolulu: The Bess Press. p. 27-28.

Bornhorst H. L., and Rauch, F. D. 2003. Ornamentals and flowers native Hawaiian plants for

landscaping, Conservation, and Reforestation. Available:

www.ctahr.hawaii.edu/hawnprop/plants/jac-oval.htm [January 7, 2019 ].

da Rocha Corea, L., Trolies J., Mastroberti A.A., Mariath J.E.A. and Fett-Neto A.G. 2011.

Distinct modes of adventitious rooting in Arabidopsis thaliana. Plant Biology 14:100-108

de Souza, E. H., Souza, F.D., Costa, M.A.P., Costa Jr., D.S, dos Santos-Serejo, J.A., Amorim, E

and Ledo C.A. 2011. Genetic variation of the Ananas genus with ornamental potential.

Genetic Resources and Crop Evolution. Vol. 59 Issue 57, pp 1357-1376

Friend A.L., Coleman M.D and Isebrands, J.G. 1994. Carbon allocation to root and shoot sytems

of wood plants. International Symposium on the Biology of Adventious Root Formation.

Plenum Press. New York, 245-274.

Hamilton C.J., Emino, E.R. and Bartuska, C.A. 2002. The influence of cutting size, leaf area and

shipping on Coleus cutting quality parameters including rooting. Proceedings of the Florida

State Horticultural Society. 115: 134-136

Hartmann, H.T. and Kester, D.E. 1983. Plant Propagation: Principles and Practices. 8th edition.

Prentice Hall, Inc., Englewood Cliffs, NJ. 725 p

Hartmann H. T., Kester D. E., Davies J.F.T. and Geneve R.L. 1997. Plant Propagation:

Principles and Practices. pp.239-391:770. Prentice-Hall, London, UK. 6th edition. ISBN-10:

0-13- 206103-1. ISBN-13: 978-0-13-206103-2.

Hartmann H.T., Kester, D.E., Davies Jr, F.T. and Geneve, R.L. 2002. Plant Propagation:

Principles and Practices. 7th ed. New Jersey, Prentice Hall. 880p.

Hooper V.H, Endter-Wada J., and Johnson, C.W. 2016. Theory and practice related to native

plants: A case study of Utah landscape professionals. Landscape Journal 27 (1):127-141.

https://dlnr.hawaii.gov/occl/files/2013/08/001-salt-tolerance.pdf

74

Husson F., Josse, J., Le, S. and Mazet, J. 2018. FactoMineR: Multivariate Exploratory Analysis

and Data Mining. R package version 1.41.

Available:https://cran.rproject.org/web/packages/FactoMineR/FactoMineR.pdf. [Accessed

April 1, 2019]

Jarvis B.C., 1986. Endogenous control of adventitious rooting in non-woody species. In:

Jackson, M.B. (ed.), New Root Formation in Plant and Cuttings, Martinus Nijhoff, pp. 191–

222.

KaiserC., and Ernst, M. 2014. Hanging baskets. Center for Crop Diversification, 1–4. Available:

www.uky.edu/Ag/CCD/introsheets/hangingbaskets.pdf [Accessed January 7, 2019].

Kessler R., Jr. 2007. Flowering hanging baskets. commercial greenhouse production. Auburn

University. Available: http://www.ag.auburn.edu/landscape/Hangbask.htm. [Accessed

January 7. 2019].

Kassambara A. 2018. ggpubr: 'ggplot2' based publication ready plots, version R package version

0.1.7. website: https://CRAN.R-project.org/package=ggpubr.

Kassambara A., and F. Mundt. 2017. Factoextra: Extract and visualize the results of multivariate

data analyses. R package version 1.0.5. Available: https://CRAN.R-

project.org/package=factoextra. [Accessed April 1, 2019]

Lerner, R. 2002. Hanging baskets. West Lafayette,IN: Department of Horticulture, Purdue

University Cooperative Extension. Available: https://ag.purdue.edu/hla/pubs/HO/HO-

126.pdf. [Accessed January 7, 2018]

Lilleeng-Rosenberger, K. 1998. Propagation techniques for native Hawaiian plants. Newsletter

of the Hawaiian Botanical Society 37 (2):33-35. Available:

https://www.ctahr.hawaii.edu/hawnprop/plants/jac-oval.htm [January 7, 2019].

Maria, C., and Bona, D. 2010. Influence of leaf retention on cutting propagation of Lavandula

dentata L ., Revista Ceres 57(4):526–529.

Namoff, S., Luke, Q., Jimenez, F., Veloz, A., Lewis, C., Sosa, V., Maunder, M., Francisco-

Ortega, J. 2009. Phylogenetic analyses of nucleotide sequences confirm a unique plant

intercontinental disjunction between tropical Africa, the Caribbean, and the Hawaiian

Islands. Journal Plant Research 123:57-65

Peres-Neto P.R., Jackson, D.A and Somers, K.M. 2010. Giving meaningful interpretation to

ordination axes; assessing loading significance in Principal Component Analysis. Ecology

84 (9): 2347-2363

Richman M. B. 1988. A cautionary note concerning a commonly applied eigenanalysis

procedure. Tellus 40B:50-58.

http://www.ag.auburn.edu/landscape/Hangbask.htm

https://cran.r-project.org/package=ggpubr

https://cran.r-project.org/package=factoextra

https://cran.r-project.org/package=factoextra

75

Ricordi A. H., Kaufman A. J., Cox, L. J., Criley, R. and Cheah, K. T. 2014. Going native in

Hawaii: opportunities and barriers for using native plant material by landscape architects.

Landscape Journal, 289.

Ruchala, S. L. 2002. Propagation of several native ornamental plants. MS Thesis, The Univ. of

Maine.,ME. Available:

https://digitalcommons.library.umaine.edu/cgi/viewcontent.cgi?referer=https://www.google.

com/&httpsredir=1&article=1454&context=etd [Accessed January 7, 2019]

Sansberro, P., Filip, R., Lo, P., Luna, C., & Mroginski, L. 2005. Effect of leaf retention and

flavonoids on rooting of Ilex paraguariensis cuttings, Scientia Horticulturae 103:479–488.

Shay B. K. R., and Drake D. R. 2018. Pollination biology of the Hawaiian coastal vine

Jacquemontia sandwicensis (Convolvulaceae ). Pacific Science, 72(4):

State of Hawaii. 2015. Act 233 Relating to Hawaiian plants. Available:

https://www.capitol.hawaii.gov/session2015/bills/GM1342_.PDF [Accessed May 6,

2019]

Starman, T. and Eixmann, K. 2006. Growing hanging baskets. Texas A and M University.

Available:aggie-horticulture.tamu.edu/floriculture/hanging-basket/index.html [Accessed

May 6, 2018]

Tamimi, L.N. 1996. The use of native Hawaiian Plants by landscape architects in Hawaii. MS

Thesis. Virginia Tech. Available: https://vtechworks.lib.vt.edu/handle/10919/31801

[Accessed January 7, 2019]

Tombesi S., Pallioti A., Poni S. and Farineli D. 2015. Influence of light and shoot development

stage on leaf photosynthesis and carbohydrate status during the adventitious root

formation in cuttings of Corylus avellana L. Frontiers of Plant Science 6 (973):1-14

Wagner W., Herbst, D., and Sohmer, S.H. 1999. Manual of the flowering plants of Hawaii.

University of Hawaii Press and Bishop Museum Press, Honolulu

Wickham, H. 2016. ggplot2: Elegant graphics for data analysis.

Available: https://cran.rproject.org/web/packages/ggplot2/index.html [Accessed April 1,

2019

Wickham, H. 2016. plyr: Tools for splitting, applying and combining data. R package version

1.8.4. Available: https://cran.r-project.org/web/packages/plyr/plyr.pdf [Accessed April 1,

2019]

https://digitalcommons.library.umaine.edu/cgi/viewcontent.cgi?referer=https://www.google.com/&httpsredir=1&article=1454&context=etd

https://digitalcommons.library.umaine.edu/cgi/viewcontent.cgi?referer=https://www.google.com/&httpsredir=1&article=1454&context=etd

https://aggie-horticulture.tamu.edu/floriculture/hanging-basket/index.html

https://vtechworks.lib.vt.edu/handle/10919/31801

https://cran.r-project.org/web/packages/ggplot2/index.html

https://cran.r-project.org/web/packages/plyr/plyr.pdf

76

Wickham, H., Francois, R., Henry, L and Muller, K. 2019. dplyr: A grammar of data

manipulation. R package version 0.8.0.1. Available:

https://cran.rproject.org/web/packages/dplyr/index.html [Accessed April 1, 2019]

Zohary, D. 2001. Domestication of crop plants. Encyclopedia of Biodiversity. pp 217-227

https://cran.rproject.org/web/packages/dplyr/index.html

CHARACTERIZATION AND PROPAGATION OF PA'UOHI'IAKA ......Agriculture NEWGERMPLASM Grant for providing...

Documents

Transcript of CHARACTERIZATION AND PROPAGATION OF PA'UOHI'IAKA ......Agriculture NEWGERMPLASM Grant for providing...